AI Paper Atlas

Research map / Adversarial robustness and security / Certified and empirical robustness

Intrinsic Explanation of Random Subspace Method for Enhanced Security Applications

Yanting Wang, Jinyuan Jia

ICLR 2026 · Adversarial robustness and security › Certified and empirical robustness · approach: Theory of adversarial risk

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use EnsembleSHAP instead of Shapley valuerandom subspace ensemble (BERT) · IMDb (no attack, delete 10%)faithfulness (label flip rate): 0.300 -> 0.600Improved
use EnsembleSHAP instead of Shapley valuerandom subspace ensemble (BERT) · IMDb (adversarial attack, delete 10%)faithfulness (label flip rate): 0.845 -> 0.980Improved
use EnsembleSHAP instead of Shapley valuerandom subspace ensemble (BERT) · AG-news (backdoor attack, delete 10%)faithfulness (label flip rate): 0.725 -> 0.735No clear effect
use EnsembleSHAP instead of LIMErandom subspace ensemble (BERT) · SST-2 (no attack, delete 10%)faithfulness (label flip rate): 0.125 -> 0.365Improved
use EnsembleSHAP instead of Shapley valueRA-LLM (Vicuna-7B) · DAN jailbreak (delete 10%)faithfulness (label flip rate): 0.33 -> 0.85Improved
use EnsembleSHAP instead of LIMERA-LLM (Vicuna-7B) · GCG jailbreak (delete 10%)faithfulness (label flip rate): 0.15 -> 0.15No clear effect
use EnsembleSHAP instead of Shapley valuerandom subspace ensemble (BERT) · IMDb (backdoor attack, top-5 words)key-word recall: 0.491 -> 0.892Improved
use EnsembleSHAP instead of Shapley valuerandom subspace ensemble (BERT) · SST-2 (adversarial attack, top-5 words)key-word precision: 0.361 -> 0.378No clear effect
use EnsembleSHAP instead of LIMERA-LLM (Vicuna-7B) · AutoDAN jailbreak (top-10 words)key-word precision: 0.335 -> 0.434Improved
use EnsembleSHAP instead of Shapley valueRA-LLM (Vicuna-7B) · DAN jailbreak (top-10 words)key-word recall: 0.289 -> 0.287No clear effect
increase number of sub-sampled inputs N (ablation)EnsembleSHAP · text classificationfaithfulness and key-word predictionImproved
increase dropping rate to a very large value (0.9) (ablation)EnsembleSHAP · text classificationfaithfulnessHurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy